Abstract
Introduction. With the expansion of single embryo transfer practices, selecting the most competent embryo is becoming critical for achieving high reproductive outcomes. Morphological assessment, including the Gardner classification, remains a widely used tool; however, its prognostic value remains limited and variable. Time-lapse imaging technologyallows for recording embryo morphokinetic parameters and developing implantation prediction algorithms based on the timing of key developmental stages. Research shows the influence of a number of factors — woman’s age, body weight, smoking, and endometriosis — on morphokinetic parameters. However, the advantages of TLM over traditional culture remain unclear. Nevertheless, the use of time-lapse imaging can improve the accuracy of embryo selection, including the selection of candidates for transfer or PGT-A, which is particularly important in patients of advanced reproductive age. Objective. To identify prognostic factors for a positive IVF outcome based on classification tree analysis. Materials and methods. Data from 431 IVF cycles were analyzed. A positive outcome was defined as clinical pregnancy, and a negative outcome was defined as its absence. The CHAID method was used to identify significant predictors. The analysis included patient age and morphokinetic parameters of embryo development by day 5 of culture. Results. The key discriminatory factor was patient age (≤ 35 years) (p = 0.004). In this group, the success rate was 47.8 %, compared to 31.0 % in patients over 35 years of age. Among women ≤ 35 years of age, the most significant factor was the morphokinetic assessment score (AI score on day 5, p = 0.001): with a score > 9.4, the probability of a successful outcome reached 72.5 %, while with a score ≤ 5.1, it was only 30.3 %. A similar pattern persisted in patients over 35 years of age (p < 0.001), but the success rate with high scores decreased to 46.5 %. Conclusion. Patient age and embryo morphokinetic assessment are independent predictors of IVF success. Using a classification tree allows for personalized outcome prognosis and optimized embryo transfer strategies.